Modify the masses

.github/workflows/tests.yml

@@ -15,7 +15,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ['3.10', '3.11']
+        python-version: ['3.10', '3.11', '3.12']
 
     steps:
     - uses: actions/checkout@v2

DOCUMENTATION.md

@@ -0,0 +1,231 @@
+# Documentation
+
+## Install
+
+To install this package, you need a running Python 3 installation (Python >= 3.10 recommended), and
+
+```bash
+pip3 install git+https://github.com/pierre-24/phonopy-vibspec.git
+```
+
+Note: as this script install programs, you might need to add their location (such as `$HOME/.local/bin`, if you use `--user`) to your `$PATH`, if any.
+
+## Theory
+
+For an intro to phonon calculations, see [the VASP documentation](https://www.vasp.at/wiki/index.php/Phonons:_Theory). 
+
+Long story short, in order to get a spectra, one needs to:
+
+1. Compute the dynamic matrix (also called mass-weighted hessian) and diagonalize it, which will provide the frequencies and modes, and
+2. Compute the Born charge in order to compute the IR intensities and/or their derivatives to get the Raman intensities.
+
+## Usage
+
+### Frequencies and normal modes
+
+To get the force constant, you need to use [`Phonopy`](https://phonopy.github.io/phonopy/index.html) (which is installed since it is a dependency of this package) as usual.
+
+On the one hand, if you can use [DFPT](https://phonopy.github.io/phonopy/vasp-dfpt.html#vasp-dfpt-interface) (which is recommended), the procedure is the following:
+
+```bash
+# 1. Create POSCAR of supercell:
+phonopy -d --dim="1 1 1" -c unitcell.vasp 
+# Note: use preferentially a larger cell!
+
+# 2. cleanup
+rm POSCAR-*
+mv SPOSCAR POSCAR
+
+# 3. Run VASP using `IBRION` (see below)
+vasp_std
+
+# 4. Extract force constants (a `force_constants.hdf5` file is created)
+phonopy --hdf5 --fc vasprun.xml
+```
+
+For step 3, the [`IBRION`](https://www.vasp.at/wiki/index.php/IBRION) keyword should be set to 5, 6, 7, or 8 (more info [there](https://www.vasp.at/wiki/index.php/IBRION#Computing_the_phonon_modes)).
+When using numerical differentiation (`IBRION=5` or `IBRION=6`), convergence criterion should be stricter.
+For example:
+
+```text
+IBRION = 6      ! numerical differentiation, using symmetry
+NFREE = 2       ! central differences for the force, should be 2 or 4
+POTIM = 0.015   ! displacement, default is 0.015
+EDIFF = 1.0e-08 ! stricter criterion for energy convergence
+PREC = Accurate ! increase precision
+```
+
+You might also want to increase [`NELM`](https://www.vasp.at/wiki/index.php/NELM) since `EDIFF` is increased.
+
+On the other hand, (see [there](https://phonopy.github.io/phonopy/vasp.html)):
+
+```bash
+# 1. Create POSCAR of supercell:
+phonopy -d --dim="1 1 1" -c unitcell.vasp 
+# Note: use preferentially a larger cell
+
+# 2. Create folders for calculations
+for i in POSCAR-*; do a=${i/POSCAR/disp}; mkdir -p $a; mv $i ${a}/POSCAR; done; 
+
+# 3. Run VASP using `IBRION=-1` (see below)
+vasp_std
+
+# 4. Extract force sets
+phonopy -f disp-*/vasprun.xml
+
+# 5. Compute force constants (a `force_constants.hdf5` file is created)
+phonopy --writefc-format HDF5 --writefc
+```
+
+Again, for step 3, you need to set:
+
+```text
+EDIFF = 1.0e-08 ! stricter criterion for energy convergence
+PREC = Accurate  ! increase precision
+```
+
+### Visualisation and interpretation of the normal modes
+
+If you want, you can then create files to visualize the modes in [VESTA](http://jp-minerals.org/vesta/en/):
+
+```bash
+phonopy-vs-modes --modes="4 5 6"
+```
+
+A `modexxx.vesta` is created per mode.
+
+Furthermore, a partial analysis of normal modes is available: it estimates the percentage of translation, rotation, and vibration of normal modes.
+
+```bash
+phonopy-vs-analyze-modes
+```
+
+However, to determine the rotation, the program needs a center of rotation.
+By default, it is taken as the center of mass. You can manually set it with `-C`, *e.g.* `-C ".5 .5 .5"` for the center of the cell (this center is to be given in **fractional coordinates**).
+It is also possible to "unwrap" the cell (i.e., move atoms close together), which gives better results for single molecules.
+
+**Note:** negative vibrational contributions are sometimes reported. Keep in mind that this is an estimate.
+
+### Infrared spectrum
+
+After obtaining the dynamic matrix, the frequencies and corresponding mode, you then need to run another calculation in order to compute the born effective charges and extract them using [a utility](https://phonopy.github.io/phonopy/auxiliary-tools.html#phonopy-vasp-born) provided by Phonopy:
+
+```bash
+# 1. Run a calculation with `LEPSILON = .TRUE.` **on the unit cell**
+vasp_std
+
+# 2. Extract Born effective charges from calculations (a `BORN` file is created)
+phonopy-vasp-born vasprun.xml > BORN
+```
+
+For step 1, the `INCAR` file should use the following parameters (again, to ensure precision):
+
+```text
+LEPSILON = .TRUE. ! Compute dielectric matrix
+PREC = Accurate
+EDIFF = 1.0e-08
+```
+
+Then, you can create an IR spectrum:
+
+```bash
+# 3. Get IR spectrum
+phonopy-vs-ir -b BORN  spectrum.csv
+```
+
+The `-b` option controls the location of the `BORN` file
+
+The output CSV file contains two sections:
+
+1. a list of each normal mode, its irreducible representation, and their IR activity, and
+2. a spectrum.
+
+You can control the latter using different command line options:
+
++ `--limit 200:2000`, which create a graph between 200 and 2000 cm⁻¹;
++ `--each=1`, the interval between each point (in cm⁻¹);
++ `--linewidth=5`, the linewidth of the Lorentzian (in cm⁻¹);
+
+It is also possible to compute spectra at other `q` points in the Brillouin zone:
+
+```bash
+phonopy-vs-ir -q="0.5 0 0" spectrum_0.5.csv
+```
+
+Note that Phonopy is generally not able to assign symmetry labels in that case.
+
+## Raman spectrum
+
+The procedure is more complex, since one needs the derivatives of the BORN charge with respect to polarizability (*i.e.*, the polarizability):
+
+```bash
+# 1. Get displaced geometries
+phonopy-vs-prepare-raman
+
+# 2. Create folders for calculations
+for i in dielec-*.vasp; do a=$(i%.vasp); mkdir -p $a; cd $a; ln -s ../$i POSCAR; cd ..; done; 
+
+# 3. Run calculations with `LEPSILON = .TRUE.` for each displaced geometry
+for i in dielec-*; do cd $i; vasp_std; cd ..; done
+
+# 4. Collect dielectric constants
+phonopy-vs-gather-raman dielec-*/vasprun.xml
+
+# 4. Get Raman spectrum
+phonopy-vs-raman spectrum.csv
+```
+
+The resulting output contains the same sections as with IR (except it gives raman activities), and can be controlled using the same command line options.
+
+## Contribute
+
+Contributions, either with [issues](https://github.com/pierre-24/phonopy-vibspec/issues) or [pull requests](https://github.com/pierre-24/phonopy-vibspec/pulls) are welcomed.
+
+### Install
+
+If you want to contribute, this is the usual deal: 
+start by [forking](https://guides.github.com/activities/forking/), then clone your fork and use the following install procedures instead.
+
+```bash
+cd phonopy-vibspec
+
+# definitely recommended in this case: use a virtualenv!
+python -m venv virtualenv
+source venv/bin/activate
+
+# install also dev dependencies
+make install
+```
+
+### Tips to contribute
+
++ A good place to start is the [list of issues](https://github.com/pierre-24/phonopy-vibspec/issues).
+  In fact, it is easier if you start by filling an issue, and if you want to work on it, says so there, so that everyone knows that the issue is handled.
+
++ Don't forget to work on a separate branch.
+  Since this project follows the [git flow](http://nvie.com/posts/a-successful-git-branching-model/), you should base your branch on `main`, not work in it directly:
+
+    ```bash
+    git checkout -b new_branch origin/main
+    ```
+
++ Don't forget to regularly run the linting and tests:
+
+    ```bash
+    make lint
+    make test
+    ```
+
+    Indeed, the code follows the [PEP-8 style recommendations](http://legacy.python.org/dev/peps/pep-0008/), checked by [`flake8`](https://flake8.pycqa.org/en/latest/).
+    Having an extensive test suite is also a good idea to prevent regressions.
+
++ Pull requests should be unitary, and include unit test(s) and documentation if needed. 
+  The test suite and lint must succeed for the merge request to be accepted.
+
+## Who?
+
+My name is [Pierre Beaujean](https://pierrebeaujean.net), and I have a Ph.D. in quantum chemistry from the [University of Namur](https://unamur.be) (Belgium).
+I'm the main (and only) developer of this project, used in our lab.
+I use this in the frame of my post-doctoral research in order to study batteries and solid electrolyte interphrase, and I developed this project to ease my life.
+
+Note: due to my (quantum) chemistry background, we may speak of similar things using a different vocabulary.

README.md

@@ -9,73 +9,6 @@ Then, calculation(s) of the dielectric matrix provide the IR and Raman intensiti
 Note: this is actually a simpler (and packaged!) version of [`Phonopy-Spectroscopy`](https://github.com/skelton-group/Phonopy-Spectroscopy). The main difference with the latter is that this package does not include phonon line widths (and thus does not require `phono3py`).
 If you are interested in that (or polarized Raman), use their code instead :)
 
-## Install
+## Installation, usage, & contributions
 
-To install this package, you need a running Python 3 installation (Python >= 3.10 recommended), and
-
-```bash
-pip3 install git+https://github.com/pierre-24/phonopy-vibspec.git
-```
-
-Note: as this script install programs, you might need to add their location (such as `$HOME/.local/bin`, if you use `--user`) to your `$PATH`, if any.
-
-## Usage
-
-Common procedure:
-```bash
-# 1. Create POSCAR of supercell:
-# (from https://phonopy.github.io/phonopy/vasp-dfpt.html#vasp-dfpt-interface)
-phonopy -d --dim="1 1 1" -c unitcell.vasp 
-# Note: use preferentially a larger cell
-
-# 2. cleanup
-rm POSCAR-*
-mv SPOSCAR POSCAR
-
-# 3. Run VASP using `IBRION=8` or `IBRION=6` with appropriate `POTIM`
-
-# 4. Extract force constants (a `force_constants.hdf5` file is created)
-phonopy --hdf5 --fc vasprun.xml
-```
-
-Create files to vizualize the modes in [VESTA](http://jp-minerals.org/vesta/en/):
-
-```bash
-phonpy-vs-modes --modes="4 5 6"
-```
-
-For infrared:
-```bash
-# 1. Run a calculation with `LEPSILON = .TRUE.` **on the unit cell**
-
-# 2. Extract Born effective charges from calculations
-phonopy-vasp-born vasprun.xml > BORN
-
-# 3. Get IR spectrum
-phonopy-vs-ir spectrum.csv
-```
-
-For Raman:
-```bash
-# 1. Get displaced geometries
-phonopy-vs-prepare-raman
-
-# 2. Create folders for calculations
-for i in dielec-*.vasp; do a=$(i%.vasp); mkdir -p $a; cd $a; ln -s ../$i POSCAR; cd ..; done; 
-
-# 3. Run calculations with `LEPSILON = .TRUE.` for each displaced geometry
-
-# 4. Collect dielectric constants
-phonopy-vs-gather-raman dielec-*/vasprun.xml
-
-# 4. Get Raman spectrum
-phonopy-vs-raman spectrum.csv
-```
-
-## Who?
-
-My name is [Pierre Beaujean](https://pierrebeaujean.net), and I have a Ph.D. in quantum chemistry from the [University of Namur](https://unamur.be) (Belgium).
-I'm the main (and only) developer of this project, used in our lab.
-I use this in the frame of my post-doctoral research in order to study batteries and solid electrolyte interphrase, and I developed this project to ease my life.
-
-Note: due to my (quantum) chemistry background, we may speak of similar things using a different vocabulary.
+See [DOCUMENTATION.md](DOCUMENTATION.md).

phonopy_vibspec/__init__.py

@@ -11,8 +11,59 @@
 __email__ = 'pierre.beaujean@unamur.be'
 __status__ = 'Development'
 
+from typing import Optional, Set
 
 # logging
 logging.basicConfig(level=logging.WARNING)
 logger = logging.getLogger(__name__)
 logger.setLevel(os.environ.get('LOGLEVEL', 'WARNING').upper())
+
+
+class GetListWithinBounds:
+    """
+    Get a list of integer within an interval `[min, max]`.
+    Open intervals are possible by setting `None`.
+    """
+    def __init__(self, min_index: Optional[int] = None, max_index: Optional[int] = None):
+        self.min_index = min_index
+        self.max_index = max_index
+
+    def _check_or_raise(self, n: int):
+        """Check that number fall in interval"""
+        if self.min_index is not None and n < self.min_index:
+            raise ValueError('{} should be larger than {}'.format(n, self.min_index))
+        if self.max_index is not None and n > self.max_index:
+            raise ValueError('{} should be smaller than {}'.format(n, self.max_index))
+
+    def __call__(self, inp: str) -> Set[int]:
+        """
+        Get a list of atom indices.
+        Ranges (i.e., `1-3` = `[0, 1, 2]`) and final wildcard (i.e., 2-*` = `[2, 3 ...]`, only if closed interval)
+        are supported.
+        """
+        lst = set()
+
+        for x in inp.split():
+            if '-' in x:
+                chunks = x.split('-')
+                if len(chunks) != 2:
+                    raise ValueError('{} should contain 2 elements'.format(x))
+
+                b = int(chunks[0])
+                self._check_or_raise(b)
+
+                if chunks[1] == '*':
+                    if self.max_index is None:
+                        raise ValueError('cannot use `*` with an open interval')
+                    e = self.max_index
+                else:
+                    e = int(chunks[1])
+                    self._check_or_raise(e)
+
+                lst |= set(range(b, e + 1))
+            else:
+                n = int(x)
+                self._check_or_raise(n)
+                lst.add(int(x))
+
+        return lst